Best microscope for wafer viewing

[EEVblog]

As the title says, what's the best microscope for viewing and more importantly video recording detail on silicon wafers and other really tiny detail stuff with excellent image quality?
I'm thinking something like this coaxial light monocular scope perhaps with of course suitable video camera on top:
http://www.ebay.com.au/itm/AmScope-H800-CL-Inspection-Zoom-Monocular-Microscope-w-Coaxial-Light-/400427943454?pt=LH_DefaultDomain_0&hash=item5d3b5d861e
Field of view 2.4mm @ 1920x1080 HD gives 1.25um resolution per pixel. if the optics are good enough, could even use a 4K camera.
Thoughts and suggestions?

[mikeselectricstuff]

For die details you're well into "proper" microscope territory - 50-1000x total mag (objective x eyepiece)

You want something that has a beamsplitter, so it fires light through the objective - sometimes called a metallurgical microscope - as with high magnification the working distance is vanishingly short - virtually touching, so there's no room for external light, and as the subject is opaque you can't fire light through the subject
x100 objectives typically need oil between objective and subject.

[Rupunzell]

Suggest the industry standard for a long time. Leitz Ergolux. These are designed for wafer inspection using bright field & dark field illumination with infinity corrected objectives and ICR prism option can be had if needed. The standard objectives for these are NPL, the higher quality objectives are Fluotar which are "semi APO". For camera imaging, make sure to get the ergo trinocular tube specifically designed for optical viewing and a camera port.

They look like this:
http://www.ebay.com/itm/Leitz-ergolux-microscope-/291401830681?pt=LH_DefaultDomain_0&hash=item43d8e71119

Ergolux in good condition with a complete set of proper optics and lighting can be purchased for $1000 or less USD depending on condition optics and all that. Shipping will be expensive due to it's extremely fragile nature and weight.. nearly 50 pounds of high precision German made optical instrument. Made in China is not the way to go here.

The alternatives are a good metallurgical microscope with bright field / dark field made by Zeiss, Nikon, Olympus, Reichert, Zeiss Jena or Leica Micro Systems. These often do not have a large enough stage with the stability and movement range required for wafer work. These do meet the optical and lighting requirements and are adaptable for camera set ups.

IMO, it is a waste of time and all related to use anything less for this kind of work. Like all high quality instrumentation, they are designed and intended to be serviced and stay in service for a very long time with proper ergonomics and user satisfaction.

 
Bernice

As the title says, what's the best microscope for viewing and more importantly video recording detail on silicon wafers and other really tiny detail stuff with excellent image quality?
I'm thinking something like this coaxial light monocular scope perhaps with of course suitable video camera on top:
http://www.ebay.com.au/itm/AmScope-H800-CL-Inspection-Zoom-Monocular-Microscope-w-Coaxial-Light-/400427943454?pt=LH_DefaultDomain_0&hash=item5d3b5d861e
Field of view 2.4mm @ 1920x1080 HD gives 1.25um resolution per pixel. if the optics are good enough, could even use a 4K camera.
Thoughts and suggestions?

[Rupunzell]

Immersion oil is nto typically used for this kind of microscope work. A 100X objective + 10X eye piece yields 1000X magnification which is enough for wafer work. Oil immersion objectives are common for transmitted biological microscope work as the transition from glass slide to air to objective's first element becomes a significant resolution limiter due to difference in refractive index. Oil immersion helps to reduce this loss.

What matter in microscope optics is resolution, contrast, color correction and image distortion. Basic magnification is easy to get, resolution, contrast, color correction with low image distortion is difficult. Far more difficult is achieving all four with a good illumination system. 


Bernice

For die details you're well into "proper" microscope territory - 50-1000x total mag (objective x eyepiece)

x100 objectives typically need oil between objective and subject.

[coppice]

Modern wafers only look interesting when you use an electron microscope.

[EEVblog]

Suggest the industry standard for a long time. Leitz Ergolux. These are designed for wafer inspection using bright field & dark field illumination with infinity corrected objectives and ICR prism option can be had if needed.

ICR Prism?
Does anyone have a link to a manual or brochure for these Leitz Ergolux?

[EEVblog]

This Amscope one looks suitable too:
http://www.ebay.com.au/itm/40x-800x-Metallurgical-Microscope-w-Polarizing-Features-5MP-Camera-/141598858149
But always, those USB cameras are pretty crap at video, but could plug in a real camera.

[EEVblog]

Ergolux in good condition with a complete set of proper optics and lighting can be purchased for $1000 or less USD depending on condition optics and all that. Shipping will be expensive due to it's extremely fragile nature and weight.. nearly 50 pounds of high precision German made optical instrument.

Yup, shipping will be the killer on that for sure.

[SeanB]

Only $200.............

[rs20]

This Amscope one looks suitable too:
http://www.ebay.com.au/itm/40x-800x-Metallurgical-Microscope-w-Polarizing-Features-5MP-Camera-/141598858149
But always, those USB cameras are pretty crap at video, but could plug in a real camera.

That one looks far superior to me, it appears to use real, standard objectives, including a 50x one. This is important, because you're getting quite a lot of microscope for your money there, which means that the included objectives might not be A+, Full HD, distortion-free ultra-objectives. So just like you're suspecting that you might want to swap in your own camera, you may wish to (ultimately, if you decide you want to) buy an objective specific to your needs (just like a photographer chooses a lens specific to his/her needs -- this analogy is very apt, a lens for an SLR camera is technically termed the camera's objective lens, they are one and the same thing*). The more expensive an objective, the less the aberration, the higher the effective resolution, and the lower the distortion -- just like SLR lenses.

* In fact, I made a microscope-objective-to-SLR adapter a while back. A standard microscope thread on one end, a Nikon lens mount on the other, and 160mm of plain old air in between. Using some objectives that I got from surplus shed optics (made in india, 4 for $20 IIRC), I got some lovely, if not full-HD, photos of the TFT transistors in a TFT monitor (had to take the monitor apart to see it; the working distance is so shallow that the front glass on an LCD is too much!). It was a massive pain to use because I had no fine adjustment knob, which is why I'm not suggesting you use the same approach. I only tell this little story as a way to make it crystal clear that a microscope is a fairly dumb, but precisely moveable tube. The camera and the objective are the stars of the show, so make sure they're both interchangeable.

[EEVblog]

That one looks far superior to me, it appears to use real, standard objectives, including a 50x one. This is important, because you're getting quite a lot of microscope for your money there, which means that the included objectives might not be A+, Full HD, distortion-free ultra-objectives.

Amscope have a very good rep on the forum it seems, at least for their lower magnification stereo optics.
Claim the use the same production line as the big boys, not sure in what respect if that is true.

[mikeselectricstuff]

Oh, and it's is pretty much essential to have an x/y stage

[mikeselectricstuff]

Has anyone found a site that explains the various terms and jargon of microscopes in plain english ?

[coppice]

Has anyone found a site that explains the various terms and jargon of microscopes in plain english ?

If you go to the web site of a maker of many types of microscope, they can't even properly explain what makes each model special.

[EEVblog]

Has anyone found a site that explains the various terms and jargon of microscopes in plain english ?

Nope!
Doesn't even seem to be consistent across manufacturers from what I can gather.
It seems you are expected to be "in the know" about all aspects of what everything does and why and what for from each manufacturer.
There seems to be about a dozen different types of compound microscope types, and just the lighting systems and the optics and paths and technology used for that alone seem to confuse the various types of practical uses with the actual technical aspect of how the lighting itself works.
Crazy stuff.

[EEVblog]

Oh, and it's is pretty much essential to have an x/y stage

Yes, that Amscope one seems to have a bottom access vernier for that, but it doesn't tell you it's actually X and Y, or if it's even a precision vernier etc.
I have been looking at various 3rd party x/y stages just in case more precision is needed.
Usually they have course and fine verniers for each axis, so not sure how the amscope one works.

[KJDS]

What resolution do you need? Seeing details of a 10nm RAM will require quite a lot more than an old transistor on a process measured in microns.

I did some work with some power diodes that I could all the detail without my glasses on, but the smallest lateral dimension was 3mm.

[EEVblog]

Yes, that Amscope one seems to have a bottom access vernier for that, but it doesn't tell you it's actually X and Y, or if it's even a precision vernier etc.
I have been looking at various 3rd party x/y stages just in case more precision is needed.
Usually they have course and fine verniers for each axis, so not sure how the amscope one works.

On second look is has this spec:

Stage Dimensions: 7-0"x 5-3/4" (180x145mm)
Stage Traveling Range: 1-3/16"x2-3/4" (30x70mm)
Vernier Division: 0.004" (0.1mm)

So I assume the vernier they are referring to is the stage and not the focus, which by implication is both X and Y.
If that is true then that sounds good enough.

[EEVblog]

What resolution do you need?

Don't know exactly, I've just wanted a good quality microscope to make good quality videos with for a long time.
I already have my video camera macro lens that does a half decent job down to 13mm field of view. And my Tagarno goes down to 10mm field of view with insane working distance. Both of those are great for general purpose teardowns.
But I'd like to have something that does say 1mm field of view for general silicon dies and other objects for teardowns etc.
A 1mm field of view for example will give 0.5um pixel resolution on a full HD video. That should be plenty for general die structure and other small object inspection.

[EEVblog]

But I'd like to have something that does say 1mm field of view for general silicon dies and other objects for teardowns etc.

Speaking of which, FOV calculations for the Amscope one are:
The WF16X eyepieces are 11mm FOV
For the objectives supplied:
x5: FOV = 11mm / 5 = 2.2mm
x10 FOV = 11mm / 10 = 1.1mm
x50: FOV = 11mm / 50 = 220um

I'm going to assume the x50 is not going to be that good an optical quality (and likely extreme working distance limitations, i.e. practically touching, so flat objects only). So the x10 objective should be quite usable for starters.
And I can always buy a top quality x20 or even x50 objective.
Although Amscope claim the objectives are plan objectives, so should be flat across most the visible area and gives the best working distance. How much remains to be seen.

But that's optical view, not sure what that translates to into the trinocular camera port?

[rs20]

But I'd like to have something that does say 1mm field of view for general silicon dies and other objects for teardowns etc.

Speaking of which, FOV calculations for the Amscope one are:
The WF16X eyepieces are 11mm FOV
For the objectives supplied:
x5: FOV = 11mm / 5 = 2.2mm
x10 FOV = 11mm / 10 = 1.1mm
x50: FOV = 11mm / 50 = 220um

I'm going to assume the x50 is not going to be that good an optical quality (and likely extreme working distance limitations, i.e. practically touching, so flat objects only). So the x10 objective should be quite usable for starters.
And I can always buy a top quality x20 or even x50 objective.
Although Amscope claim the objectives are plan objectives, so should be flat across most the visible area and gives the best working distance. How much remains to be seen.

But that's optical view, not sure what that translates to into the trinocular camera port?

As for what the camera sees, these aren't infinity-corrected objectives*, so it's quite straightforward (once you've spent a while studying all this :-)): the eyepiece is not involved at all, the objective projects its image directly (well, via some mirrors/prisms) on the camera sensor, but 5, 10 or 50 times bigger than real life. The field of view of the supplied camera is 5.70mm(H) x 4.28mm(V), but there's a 0.5X reduction lens on it which makes it effectively 11.4mm x 8.56mm; divide that by 5x, 10x, or 50x to see the field of view that you'd get. Since your objective has a field of view of ~11mm, and the camera field of view is ~11.4mm wide, the answer is, as you would hope, almost exactly the same as the calculations you gave above. The pixel size is 3.2um, which is actually 6.4um if you factor in the reduction lens, again, divide by 5, 10 or 50 to see the "real" pixel size (remembering that each pixel is either R, G, or B, so your true resolution is considerably worse than this.)

* With infinity-corrected objectives, the image is formed by a lens in the tube; the magnification factors given by an infinity-corrected objective is stated under the assumption that the tube lens has a 200mm focal length. If it's not, you have to scale the objective magnification accordingly, a x10 infinity objective with a 100mm tube lens is really only x5.

[Rupunzell]

ICR is Leitz's version or brand name for differential interference contrast (DIC) microscopy.
http://www.microscopyu.com/tutorials/java/dic/reflectedwavefronts/

Ergolux manuals can be ordered from one of the many independent Leitz service folks. Leitz-Leica was a highly customer service oriented organization before they were taken over by Danaher (just like Tektronix, Fluke and...) after which became Leica Micro Systems. Much like what happened to Tektronix, Fluke and... post take over, the company was never the same 

Here is a short about Leitz microscopes including the Ergolux:
http://www.buntgrp.com/used_microscopes.htm

Microscopes like these can be configured to most any requirement. They are modular and have highly interchangeable parts to achieve the desired configuration.

Serious microscopes are much like serious test gear made by hewlett packard, Tek and others that were designed to the limits of physics, in the case of optical microscopes, the behavior of light, ultra precision optics manufacturing, highly specialized optical coatings and machined metal parts accurate to wave lengths of light. This is the difference between a proper high quality microscope and mass market low cost microscopes. Good optical microscopes when new are easily 5 to 6 figures of USD. Back when microscopes were mostly built to this level of quality, durability and performance, they were intended to be in service for decades, used by skilled personal who understood how to use them and care for them.
 

Bernice

[/quote]

ICR Prism?
Does anyone have a link to a manual or brochure for these Leitz Ergolux?
[/quote]

[Rupunzell]

Still a toy..

Limited lighting options, the mechanicals are rink-dinky unstable and totally unsuitable for high magnification. It appears to be transmitted light (useless for viewing wafers) and only bright field lighting, no dark field and no infinity corrected objectives. All of which are serious limitations for lighting and viewing options. Those glass slide clips are a joke as those are intended for glass microscope slides and they are a total pain in the arse to use. No decent biological transmitted light microscope today has those clips.

Consider for a moment objects being view with a microscope are typically microns in dimension. This means the related mechanical bits must have similar precision to impart any useful microscope stage movement.


Bernice

This Amscope one looks suitable too:
http://www.ebay.com.au/itm/40x-800x-Metallurgical-Microscope-w-Polarizing-Features-5MP-Camera-/141598858149
But always, those USB cameras are pretty crap at video, but could plug in a real camera.

[Rupunzell]

Spend for proper shipping cost. This is an investment for high quality instrument. $200 USD for shipping is not high or extreme considering what is being transported.


Bernice

Shipping will be expensive due to it's extremely fragile nature and weight.. nearly 50 pounds of high precision German made optical instrument.

Yup, shipping will be the killer on that for sure.
[/quote]

[Rupunzell]

Here are two.

http://www.olympusmicro.com/index.html

http://www.microscopyu.com/tutorials/


Bernice

Has anyone found a site that explains the various terms and jargon of microscopes in plain english ?